Romanoslavica XLIII

79

LINGUISTIC MARGINALIA ON SLAVIC ETHNOGENESIS

Sorin PALIGA University of Bucharest

Introduction

Not only once indeed I approached a linguistic view on the Slavic ethnogenesis (to just use a consecrated term) or the ‘Slavic making’ (if to use Curta’s formula, much referred to during the last years). Disregarding whether using the traditional formula ‘ethnogenesis’ (now – perhaps not without reason – in decay) or ‘making’, the topics for debate are of course the same: where could we possibly locate the Slavic ethnogenesis / making (be it a restricted or large area), within what time span, and on what basis? As ‘origin’, ethnic or not, has always been a philosophical or legendary question, disregarding the topic in view, I shall try a linguistic, and occasionally an interdisciplinary, view on the Slavic ethnogenesis. I stress, from the very beginning, that I do not intend to review Curta’s book, which is an archaeological approach (beyond my competence), but to point out the relevant data and conclusions of Curta and other authors. Some of Curta’s views have been advocated, at least partially, by other authors as well. Putting together the views of Godłowski or Ján Pauliny (in his remarkable Arabské správy o Slovanoch) and, with reader’s generosity, my view advocated over years (see the references), I think we may now contour a reliable base for discussion. Adding here the remarkable contribution of Aleksandar Loma presented at the 13th International Congress of Slavists in Ljubljana, august 2003, one may now have a quite large and comprehensive horizon of what we may plausibly label the Slavic ethnogenesis or, in Curta’s words, ‘the making of the Slavs’.

Romanoslavica XLIII

80

The concept of ethnos

Indeed, we should first clarify, as far as possible, the concept of ethnos. Curta is definitely right in pointing out that ethnos has had variable and interpretable connotations over time. It is customary to define ethnos as referring to a certain group of people sharing a common language, similar or identical habits over a large or restricted area, and common religious beliefs. Also, an ethnos has the conscience of its identity, and defines itself as different from other groups by at least one of these basic elements. But was this interpretation valid in all times and in all circumstances? Specifically was this definition valid or understandable with, and by, the first Slavic groups as we know them from earliest historical sources?

I repeat my regret that, at least according to my knowledge, there is no global approach to the emergence (or ‘making’) of the ethnic groups of Europe beginning, say, with the 5th century A.D. Indeed, we always speak of ethnic groups (nations or peoples in modern, post-Romantic terminology), but we do not even have a clear definition of how they emerged in history. It is customary to say that the Greeks or Romans were the creators of a European identity, but we are not able to define the ethnos Greek v. any other similar group of the antiquity. It is banal to assume that the Greeks were different, but what made them different from others? Curta used the term ‘making’ in referring to the Slavs, but I could not identify any phrase in which he may have compared the making of the Slavs to the making of other ethnic groups of those times. This is, in fact, an essential minus (so to speak) of this remarkable book: in what were the Slavs different from others? Curta offers no answer at this point. He had probably assumed that readers may easily agree on the presupposed argument that they were different in se, considering their language, habits or social behaviour. But are these assumptions so obvious?

I do not wish to bore the reader with banalities, but again – I think – we do not have a clear comparative tableau of the major ethnic realities of the first millennium A.D., even if we believe that belonging to a certain ‘nation’ is a given fact, and that any person must have an ethnic identity, and this should not have any further explanation. It is now common to discriminate a French against a German because the former speaks French and the latter speaks German. But what was the criterion 15 centuries ago? I shall try to show, hopefully even to demonstrate, that things may have been different in those times, and not only referring to the Slavs. Anticipating the conclusions, I have all the reasons to

Romanoslavica XLIII

81

believe that the first Slavic groups had no generic, or common, conscience of their origin, and that the generic concept of Slavic ethnikon gradually got contours across the following centuries, to eventually become an accepted fact in the 10th century and later.

There are various perceptions of an ethnikon even in contemporary times. English, as an example, does not have a correspondent of French ethnie, and the differences between nation and people are different in every language we may analyse. To say nothing of various denotations and connotations of nation during the periods of Nazism and Communism as a forensic analysis may complicate our approach.

With these in mind, I shall attempt to have a brief look at the Slavic ‘making’ as compared to some other parallel ‘makings’. Otherwise put, to see what is common to, and what is different from, other similar situations. The Age is generous, as we may compare a series of parallel phenomena, with their similar or different aspects. The Slavs and their ‘making’ were just a chapter among other chapters of European making. I shall try to analyse only some relevant situations.

Sclavi, Sclaveni, Sclavini; Anti; Venedi

The term Sclavus, pl. Sclavi (initially used in Byzantium) and Sclavini, Sclaveni (used in most written documents) emerged in the 6th century A.D. and is currently associated with the oldest proofs of the Slavic expansion. Other sources refer to the Anti, and even older sources refer to the Venedi (as in Tacitus’ De origine et situ Germanorum). It is often held that all three refer to the Slavic groups, even if they are chronologically discriminated and definitely had different meanings across time. Were the Venedi in Tacitus the precursors of the later Sclaveni or Sclavini? If so, how may we possibly draw a plausible contour of their evolution?

What kind of ethnikon was Sclaveni, Sclavini? The question may seem bizarre, but – as shown below – not superfluous. The term emerged in the Byzantine sources in the 6th century A.D., and rapidly spread over a vast area. A comparative analysis shows that it hardly referred to a ‘pure’ ethnikon in the modern or contemporary meaning, but to the (initially) more northern groups with whom the Byzantines began to have constant, and more and more frequent military conflicts. Curta convincingly shows that, despite a largely spread hypothesis, we may hardly speak of ‘pure’ Slavs during the 6th century, and not

Romanoslavica XLIII

82

even a century later. But who were the ‘pure Slav’ in those times? And what did Sclaveni mean? A comparative look at the documents leads to the following contour:

1. The Sclaveni (initially) were of northern origin (as compared to the Byzantines, i.e. they came across the Danube); later on, they began to settle in South Danubian regions as well, but – even so – they were located north from the Byzantines, as the Empire shrank to south.

2. They were NON-Christian (a crucial detail for those times), and were important (but not unique) representatives of the Barbaricum.

3. They spoke a language, or rather languages or idioms, more or less related, perhaps often without any linguistic affinity; the Byzantines did NOT understand these languages1, and is hardly believable that the idiom – or rather in the plural, idioms – spoken by those ‘intruders’ had any relevance to them. This explains why, in some sources, there are details on recurrent misunderstandings and disagreements, which – in some cases at least – may be explained as a normal linguistic difficulty to understand each other. The linguistic barrier has always been a major impediment in mutual understanding or, in a perhaps better phrasing, has been the main reason of misunderstanding. There may be little doubt that the first contacts between the Byzantines and the new comers were marked by frequent misunderstandings as a result of linguistic barriers, of different mentalities and of a different social behaviour.

4. Militarily, they were enemies, another crucial detail, which in fact discriminated the Sclaveni against other groups of those times, e.g. against the Anti, who seemingly were linguistically related to the Sclaveni, but not enemies of the Byzantines. The dichotomy military enemy (the Sclaveni) v. military non-enemy / ally (the Anti) seems a crucial detail in those times, also reflected in the ethnikon used by the Byzantines, disregarding whether the two groups spoke similar or divergent dialects. We are rather inclined to assume that they really spoke convergent, presumably mutually intelligible, idioms construed around a South Baltic and East Iranic satem structure, with not-at-all unimportant North Thracian (Dacian) elements. Beside this nucleus, at a given moment representing perhaps the majority, there for sure were various other more or less integrated

1 Perhaps some readers would have expected to write ‘this language’, but – as shown below – we are still some good time before the linguistic coagulation later known as Old Slavic or, in its literary form, Old Church Slavonic. I have not only the feeling, but hopefully also the arguments, that these groups rather spoke more or less related idioms, still not coagulated around a congruent grammatical structure.

Romanoslavica XLIII

83

ethnic groups, some of them of Indo-European origin, others – we may be sure – of a completely different origin, e.g. the Altaic groups of the Avars.

There may be infinite debates whether only some groups spoke what we may label Proto-Slavic or a kind of ‘common Slavic’, or whether only some representatives of these new groups spoke this idiom. As shown below, there are all the arguments showing that under the term Sclaveni there were various linguistic groups, some of them perhaps without any linguistic affinity at all.

In some sources there are the Anti, another ethnikon held for another Slavic group. Who were the Anti?

1. The Anti were also of northern origin, and some reliable sources locate them in North-East regions, approximately east from the modern Romania. If sources be again reliable, they were not immediate neighbours of the Byzantines (an important detail).

2. They were NON-Christian and, like the Sclaveni, were also represen-tatives of the Barbaricum.

3. They spoke a language, or rather languages/idioms, probably related to that, or rather to those, spoken by the Sclaveni. We may guess, but only relying on later realities, that the Anti and the Sclaveni spoke perhaps related dialects of the same idiom or, better, that most of them spoke such idioms. This does not mean that we may have a clear linguistic equation, but that we may plausibly surmise a certain linguistic affinity in course of coagulation.

4. Militarily, they were non-enemies, perhaps even allies, another crucial detail, which in fact discriminated the Sclaveni against the Anti. We have all the reasons to assume that the main difference consisted in the ally v. enemy character of them (i.e. Anti v. Sclaveni) as viewed from Byzantium.

Who were the Venedi? Tacitus located them east of the Germanic groups, which some linguists took for a clear proof that they must have been the Proto-Slavs. The Venedi were, unlike the Sclaveni and the Anti, a kind of legendary people, historically with oldest references regarding the location east of the Germanic groups. This ethnic name (ethnos) may possibly be closer to our modern understanding of the meaning. For sure, some people used this name, as proved by Greek venetikós, Romanian venetíc ‘non-Christian’. Finnish venäjä ‘Russian’ also speaks of its old history. This does not mean the Venedi were Proto-Slavs and indeed there is no evidence they may have been so, it just means that, if not indeed some kind of Proto-Slavs, they were later acculturated, and held for a Slavic group. They may, or may not, be a similar case like Vlakh by

Romanoslavica XLIII

84

which the East Slavs refer to Romanians, while West Slavs refer to Italians, even if – initially – they were a Celtic group, later Romanised1.

The presumed Proto-Slavic Venedi were of course different from the Venedi, Veneti who gave the name of the city of Venice. The Proto-Slavic Venedi may have been a Celtic group too, even if such a view has a major impediment: there are no proofs of Celtic influences in Proto-Slavic. If these Venedi were also Celts (as their name may suggest), then a minimal set of Celtic words should be identified in Proto-Slavic. There is no such example. Therefore, assuming that Tacitus’ spelling was more or less correct or approximated the original form, these Venedi had their legendary or semi-legendary history as proved by preservation of forms venetikós, venetíc in southern Europe, and venäjä ‘Russian’ in Finnish. As the Finns witnessed, as neighbours, the long and complex process of Slavisation, one may credit Finnish with a good proof that indeed an ethnikon Venedi was used for the people inhabiting those areas, even if their contribution to the Slavic ‘making’ proper is obscure and undecipherable.

Sclavus/ Sclavenus; Ṣaqlab (Ṣiqlab, Ṣaqlāb), pl. Ṣaqāliba; Shtip,

Shtiptar

The term sclavus, pl. sclavi, sclaveni, sclavini is indeed post-classical, and emerged in association with the new ethnic groups of the early Middle Ages. It was used not only in written documents, but – without any doubt – in colloquial Latin, as proved by Romanian șchiau, pl. șchei < Sclavus, Sclavi. It is now obsolete, and used in place-names and (rarely) personal names only. The Șcheii Brașovului (lit. ‘the Slavs of Brașov region’, in Romania) is perhaps best known. Romanian, as in other situations, is crucial in understanding the general ethnic ‘making’ of Southeast Europe. The Byzantine sources use the form Σκλάβοι, Sklavoi, when Greek β was already pronounced v. Pauliny 1999: 35 convincingly explained the evolution of this Arabic form of Byzantine origin. And on p. 37, Pauliny stresses:

Treba však povedať, že Slovania boli prvým svetovlasým európskym etnikom s bielou

pokožkou, s ktorým sa Arabi stretli. Azda preto niektori arabski spisovatelia používali pomenovanie Ṣaqāliba aj na označenie národov, ktoré žili na severe a východe Európy. [...]

1 Etymologically, Vlakh is related to Welsh, Wales.

Romanoslavica XLIII

85

Medzi Slovanov niektori autori počítali aj Nemcov (v arab. orig. Nāmǧin alebo Nāmǧīn), lebo mali bielu kožu a svetlé vlasy a žili v susedstve Slovanov.

(We should also add, that the Slavs were the first blond ethnic group with

white skin whom the Arabs met. This was perhaps the reason why some Arab writers used the term Ṣaqāliba with reference to ethnic groups living in North and East Europe as well. [...]

Some authors assumed that the Germans (in the Arabic original Nāmǧin or Nāmǧīn) were also Slavs, as they had white skin and blond hair, and also lived in the vicinity of the Slavs.)

There is not much room here to expand on Pauliny’s remarkable book. It is just sufficient, for the limited purpose of this paper, to note that the ethnikon Sclavi, Sclaveni or, in Arabic, Ṣaqāliba, was used with reference to completely different ethnic groups, having in common their look: blond hair and white skin. For our modern ‘scientific’ approach, defining an ethnic group by only referring to their look may seem unacceptable, even humorous, but it was sufficient for those times; and entirely corresponding to their immediate needs: the Slavs represented a blond group of Slaves. What is important, in certain historical periods, the Arabs also used the same term, Ṣaqlab (Ṣiqlab, Ṣaqlāb), pl. Ṣaqāliba, as referring to other blond ethnic groups, obviously having nothing in common with the Slavs, bar their look: blond hair and white skin.

The ultimate origin of Byzantine Sclavus, pl. Sclavi, Sclavini, Sclaveni is obscure. Curta says it is a ‘Byzantine construct’. Indeed so, nevertheless words rarely spring out from themselves. A linguist would rather look for its possible origin. As commonly assumed by most scholars, it seems to be a deformation of Slověninъ, pl. Slověne, the name later used by the Slavs to discriminate themselves against other ethnic groups. This happened much later, but we may assume that some ethnic groups of those times (6th century A.D.) used this form, or this proto-form, as their ethnic name. As long as both the Slovenes (slovenec, slovenski jezik) and the Slovaks (Slovák, slovenský jazyk) still preserve this name (and also the Sloveni attested among the East Slavs before the 10th century A.D.), we may assume that the most important ‘Sclavenic’ group the Byzantines first met used this form in order to discriminate themselves against others. These must have been the precursors of the Slovaks and/or Slovenes or another group using this name.

It is true that the phonetic evolution is not clear, but – in this case – we must look not for an accurate reconstruction, but rather for a possible way from

Romanoslavica XLIII

86

the real pronunciation to the form used in documents and adapted/adopted by the Byzantines. As noted above, Romanian șchiau, șchei show that the word indeed circulated at colloquial level. If some may still think that the reference form Slověninъ, pl. Slověne is quite far from Sclavus, Sclavenus, Sclavinus, where the sequence scl- in indeed difficult to explain, we may think at another origin, be it difficult to identify. To note that the Byzantines did not care for an accurate borrowing, but to approximately note and adopt/adapt a foreign form to describe a minimal discrimination: ‘the Sclaveni were our enemies’. It was entirely irrelevant to them whether this was or not an accurate transcription of any original form, it served their needs for identification and nothing more. I assume, until further counter-arguments may be invoked, that the ‘Byzantine construct’ Sclavus, Sclavenus, Sclavinus is a deformation of what may have been a proto-form of Slověninъ, pl. Slověne, in its turn derived from slovo ‘word’, and opposed to něm- ‘dumb’, therefrom němьcь ‘German’ (i.e. ‘those who are dumb = speak a language we cannot understand’)1.

I assume that the ethnikon Sclavus, Sclavi, Sclaveni was adapted from a colloquial Romance form, as proved by preservation of form șchiau, șchei in Romanian. There is little doubt that Proto-Romanian did have a form *scla-w-us > *sklya-w-us > *ški̯a-, some time later adapted and adopted by the Byzantine documents. Even if the origin of this form may be debatable or obscure, the only reasonable explanation is that Romance population (or Proto-Romanians) adapted/deformed the original form Slověninъ, pl. Slověne. We may also consider another origin but, disregarding the source, the word gleaned into colloquial East Romance, hence into the Byzantine documents beginning with the 6th century A.D.

Some time later, the Byzantine form Sclavus, Sclavenus, Sclavinus was loaned/adapted by the Arabs as Ṣaqlab (Ṣiqlab, Ṣaqlāb), pl. Ṣaqāliba ‘Sclavus, pl. Sclaveni’, and – as Ján Pauliny noted – reflected more or less the same meaning: not accurately the linguistic affinity (even if, we may assume, most of them were Slavs, or Proto-Slavs), but their social status: slaves. For the Arabs, their were blond slaves, an entirely outstanding fact for those times, even if some of them were not, for sure, Slavs, but of various others origins. If blond and having white skin, they were Ṣaqāliba.

1 Cf. Hungarian magyar and magyarázni ‘to explain, to speak clearly’. The Magyars defined themselves as the ones who speak clearly, i.e. the same language, as opposed to foreigners.

Romanoslavica XLIII

87

As long as the Sclaveni were, by definition, the enemies of the Empire, for a long time the most important enemies in that area, they were often defeated and, of course, some of them were enslaved. The Arab documents show how, and why, the association Slav – slave gradually became equivalent: the Byzantines praised themselves for defeating them, then enslaving them, so some of them were sold to the Arabs. To note also that the Ṣaqāliba at the court of khalifs may have been of any origin as long as they were – a shocking detail for the Arabs – blond. Until those times, they had never seen blond slaves: any blond slave, e.g. of north European origin, was a Ṣaqlab. Some of them were, beyond any doubt, of Germanic descent; or of any ‘blond’ origin. Unlike the Byzantines, or to a less extent, the blond hair was a discriminating factor for the Arab world. As long as both Proto-Slavic or Germanic meant the same thing to their ear (at least initially), a language they could not understand, the discriminating factor was the social status (slaves) and the ‘aesthetic’ aspect: blondness1. The military connotation of Sclavenus (enemy of the Empire) was lost in the Arab world, as it was irrelevant.

After briefly noting the situation of the forms in Slavic Slověninъ, Slověne < slovo ‘word’; Rom. șchiau, pl. șchei; Byzantine Sclavus, pl. Sclavi, Sclaveni, Sclavini; Arabic Ṣaqlab (Ṣiqlab, Ṣaqlāb), pl. Ṣaqāliba, let us try to briefly analyse some relevant forms in Albanian. Shqip, shqipë (adj.) ‘Albanian’ is the word by which the Albanians discriminate themselves against other ethnic groups; shqiptar is the noun (‘an Albanian’). It is not a rare case when a certain ethnic group uses another form than the foreigners. There are many similar examples: suomi, suomalainen ‘Finnish, a Finn’, hay, Hayastan ‘Armenian, Armenia’, euskara ‘Basque’; Deutsch is etymologically related to Dutch, but – in modern times – they refer to different nations, even if both of Germanic descent.

It is interesting to note how Albanian and Aromanian forms support, and are supported by, the others forms analysed in this context. I quote after Vătășescu 1997: 437:

șchiau drom. s.m. ‘a Slav”, Arom. șcl’eáŭ ‘a servant, a slave’; [...] Șcheia (Suceava),

Șchei (Brașov), place-names derived from șchiau. Lat. sclavus s.m.; preserved in Italian, French, Spanish, Portuguese with the meaning ‘a slave’. [...] Sclavinica, the Byzantine name of former Dardania, has been preserved in Albanian Shqinikë ‘Bulgaria’, which may be equated with

1 Even if this view may seem unacceptable to the modern perception, I may assume that – to an unexperienced ear – Norwegian or Finnish are similar, and Lithuanian a kind of Estonian dialect, or vice-versa.

Romanoslavica XLIII

88

Romanian place-names derived from sclavus. Alb. Shqa s.m. ‘a Bulgarian; an Orthodox Greek; a heretic’; also Shkla, pl. Shkle ‘a Bulgarian’; Old Geg dialect Shqeni ‘Schiavonia, Sclavonia’ < shqe pl.; Tosc is Shqeri. [V 437]

Albanian shqip must reflect the same original form sclavus; for Albanian,

the proto-form may be reconstructed as *skljab, *skljap. It is immediately related to the already quoted Romanian șchiau, șchei. It also reflects the oscillating pronunciation of post-classical b/v, which ultimately led to confusing them in Romanian: veteranus > bătrîn ‘old (man)’, and to their complete disappearance in inter-vocalic position in most cases (not all, though), as in sclavus, sclavi > șchiau, șchei. The preservation of this form in both Romanian and Albanian is, I think, the best proof that modern Albanian got its modern shape after a more northern influx which, amalgamating with the local Romanised population of Illyrian origin, ultimately led to the ‘making’ of a new ethnic group. This is in accordance with the views advocated mainly by some scholars during the last decades: the Neo-Thracian, rather than Neo-Illyrian, origin of Albanian. Late prof. I.I. Russu also advocated this view in the 1980’s: the Albanians must reflect an ethnic move to south of some non-Romanised north Danubian Thracian groups, presumably the Carpians (who indeed had an important role among the Daci Liberi). These Thracian Carpian groups, the presumed ancestors of the Albanians (maybe also in congregation with some scattered, non-Romanised southern Thracian groups, i.e. those inhabiting the Haemus heights), came together with the Sclaveni and, for the Byzantines, they were militarily similar: enemies. Linguistically, they spoke an unknown idiom, anyway not understandable by the Byzantines (incidentally, another satem idiom as most of those spoken in Central-East and South-East Europe). In other words, disregarding whether they spoke a kind of Proto-Slavic or a late form of Thracian, they were, of course, Sclaveni, as defined above:

1. Of (more) northern origin. 2. They were NON-Christian. 3. They spoke a language, or rather languages or idioms, more or less

related, even without any linguistic affinity; incidentally, the Proto-Slavic Nucleus, Thracian and West Iranic were all languages of satem character.

4. Militarily, they were enemies. Both the Proto-Slavic and North Thracian groups are defined by these

elements. And, for sure, other ethnic groups, later assimilated and lost in the neighbouring cultural and linguistic environment. Romanised Dacian (North

Romanoslavica XLIII

89

Thracian) groups represented the East Romance element, the Proto-Romanians, and other Thracian, later Romanised groups, still reflect the complex process of Romanisation and acculturation, both in north and south Danubian regions.

The analysis of the Romance and Thracian elements in Romanian and Albanian has shown that they once were neighbours. This neighbourhood was initially north Danubian, but continued in the south Danubian regions too, where the Romance element – both of Proto-Romanian and Proto-Dalmatian character – was dominant. This view also explains why the Romance elements in Albanian reflect the local, Dalmatian character, but some also their Proto-Romanian character. The dichotomy North-Danubian v. South-Danubian has always been relative: it was important during some historical periods and were totally irrelevant during other periods. Our task is to discern when and why it was so.

The Slavic Homeland

After this brief survey of the historical, archaeological and linguistic data,

the next step should be an answer to an old question: where may we possibly locate the Slavic homeland (Urheimat, pravlast)?

Curta documented that archaeology can hardly discriminate the various artefacts from the 5th to the 10th century A.D.: the same, linguistically related, groups may have shared different cultural data or, vice-versa, different, linguistically unrelated groups, may have had common symbols as reflected in these artefacts. It must have been so, and is the best proof that archaeology alone cannot trace any reasonable, reliable equation archaeological culture = linguistic affinity. Only interdisciplinary research, still rare in general, and indeed rare if referring to this area and within the time span from, say, 5th to 10th centuries A.D., may possibly lead to coherent data. Curta himself disconcerts the reader exactly on the last two pages of his book: after stating that archaeology cannot trace back any proof for the equation archaeological culture = linguistic affinity/heritage, abruptly says that, already in the 10th century A.D., the Slavs represented a vast and impressive cultural and linguistic factor from Central to East and Southeast Europe. Instead of a conclusion, Curta disconcerts his readers by an abrupt assertion, never discussed in the preceding 600 pages! If archaeologically the selected area was so chaotically organised, how could it be so abruptly well organised in the 9th–10th centuries, at least from a linguistic point of view? How can we reconstruct a solid, coherent linguistic tableau of the late two centuries of the first millennium A.D.?

Romanoslavica XLIII

90

As long as archaeology itself (and alone) cannot offer a clear answer, then we must combine its data with historical data (vague and confusing as they are) and, of course, with linguistic data, which – despite their old history – are not so easy to interpret; and some must be surely re-interpreted. The concept of a ‘pure Slavic language’ (echtslavische Sprache) descending directly from Indo-European should be abandoned; this does not mean that Proto-Slavic abruptly emerged some time after the 5th century A.D. Let us try to resume the facts.

The Slavic languages witness a satem character, with an obvious relationship with Baltic and, I would dare add, to the Thracian elements in Romanian and Albanian1. Only a good comparative linguistic analysis may show that there were complex linguistic and cultural interaction, which erroneously led to assuming that almost all the non-Latin elements of Romanian must be of Slavic origin, based on the simplistic assumption that if in Romanian and some Slavic neighbouring languages, they must be Slavic. Some of this common vocabulary of Romanian and Slavic is, I may say, of Thracian origin, and also proves the important role of some late Thracian groups in the Slavic linguistic coagulation. And some old East Romance (Proto-Romanian) elements were also borrowed in what we may label Proto-Slavic or, better, represented a component of the coagulating linguistic and cultural process, which led to what was later known as slověnština. As some Thracian elements of Romanian indeed are quite similar to the Slavic equivalents, most linguists - beginning with the 19th century – hastened to postulate ‘a massive Slavic influence in Romanian’. In some cases, the discrimination is indeed difficult, yet – with good linguistic tools – perfectly possible. It may be also less convincing for all those, perhaps the majority, who are still inclined to such an obsolete view, but with solid roots in the Romantic and Post-Romantic atmosphere of the 19th century. Given the limited purpose of this paper, I shall quote some examples only; many others were analysed in my previous studies.

– Rom. sută ‘one hundred’, long held for a Slavic loan-word, is – beyond any reasonable doubt now – of Thracian origin. Furthermore, it was early included in the numeral system of Slavic where sъto has an obvious isolated position. As I extensively wrote on this case, I shall not insist2.

1 The author is inclined to consider Albanian a Neo-Thracian, not a Neo-Illyrian idiom. 2 I would just note the critical comment of Marko Snoj in the 3rd posthumous volume of Bezlaj’s Etimološki slovar slovenskega jezika, with Marko Snoj and Meta Furlan as editors. Snoj writes (p. 318): ‘[...] Še manj utemeljeno je mnenje, po katerem je psl. *sъto izposojeno iz dak. *su(m)tə

Romanoslavica XLIII

91

– All the series of the would-be ‘oldest Slavic borrowings’, with a long acceptance among the linguists of the 19th and 20th centuries (gard, stăpîn, jupîn etc.) reflect Thracian elements, and were later used by Slavs too.

– The term trъgъ is either Thracian or Illyrian; the earliest possible attested form is in the former Illyrian area.

– The list of the Thracian elements of Romanian is now indeed long, and includes over 1,300 forms, in both vocabulary and place-names, for which see my recently published Etymological Lexicon of the Indigenous (Thracian) Elements in Romanian. Some of them indeed interfere with the Slavic elements, but a good linguistic analysis may almost always discriminate them against each other; some cases are unclear, but may be clarified later, when additional data may be gathered together.

– Sl. kъmotra reflects Rom. cumắtră < Post-classical *komatra, *kumatra, classical commater. It must be a quite early borrowing, presumably from Proto-Romanian. Some South Slavic idioms and Albanian preserve it as kuma, which is – as Skok assumed – a hypochoristical form.

Some West Iranic elements are also relevant, among these bogъ and rajь are essential as they refer to basic religious beliefs.

The common, quite limited, Slavic-Germanic forms show that a certain interference with Germanic groups cannot be ignored: Gothic hlaiba – Slavic chleba shows a quite close cohabitation; and the series represented by the administrative terms like cěsar and knędz.

The greatest bulk is yet represented by the common Baltic-Slavic elements. The list is too long to be presented here.

Therefore, and trying to sum up these data, the linguistic situation may be summarized as such:

– The most numerous elements of vocabulary are shared by Slavic and Baltic. Loma labelled this Proto-Slavic A.

< *k̑ṃtom, kar naj bi se ohranilo v rum. sútă in trak. atpn. Σουντους. Rum. sută navadno pojmujejo kot slov. izposojenko [...]’; then quotes Miklošić here, in order to give force to his feeble ‘arguments’. I would just add that a hypothesis may be ‘utemeljeno’ or ‘neutemeljeno’, never ‘manj utemeljeno’. In his subsequent Etymological Dictionary of Slovene, Dr. Snoj eliminated this comment altogether, and not even mentioned the comments of, say, Machek regarding Czech sto. I have all the arguments, some stated here, that Dr. Snoj is NOT right in his criticising me, and is NOT right in his etymological analysis of Sl. sъto. He also ignores a lot of other arguments, all pointing to the non-Slavic character of Rom. sută.

Romanoslavica XLIII

92

– Next in range are the Thracian and West Iranic elements; the issue of the Late Thracian and Early East-Romance (Proto-Romanian) elements in Early Slavic (I once used the formula Pre-Expansion-Slavic or PES) seems to be a quite new trend in the field of Slavic studies and was seemingly opened by Giuliano Bonfante in 1967, but timidly continued since then. Loma labelled West Iranic elements as Proto-Slavic B; following his line of thought, I would dare label Late Thracian elements as Proto-Slavic C.

– The Germanic elements seems to be least numerous, yet relevant just because of this detail.

– Other few elements are shared with Finno-Ugric languages, e.g. kniga – Hungarian könyv, Slavic slovo – Hung. szó, szava, even if the phonetic correspondences are quite obscure.

– Other few elements are obscure, and may reflect other origins, Altaic and ‘Eastern’ in general.

– The Celtic elements in Proto-Slavic seem to be absent. If ever identified, they must be of minor importance, and irrelevant as a whole. This is also a crucial detail.

Trying to put together all these linguistic data, and comparing them with the most relevant archaeological and historical elements, my reconstruction is the following:

– The Proto-Slavic nucleus was represented by a more southern branch of the so-called Balto-Slavic family of the Indo-European family. This would be the PRIMARY HOMELAND, which must have been north of the Northern Thracians (the Costobocae), west of Germanic and east of West Iranic. This roughly corresponds with southeast Poland and southwest Ukraine. Across time, this nucleus was moved slightly to west (Polish school, but not Godłowski) or to east (the Russian-Ukrainian school).

– Some time after the 4th century A.D., more probably at the end of the 5th and beginning of the 6th centuries A.D., these groups began to move south and closely interfered with the northernmost North Thracian and North-East Thracian groups, some not yet Romanised, the Costoboces and Carpi; they also interfered with the Romanised Dacians (Proto-Romanians).

– In this move, the Proto-Slavic nucleus also interfered with West Iranic groups, and perhaps assimilated some Uralic (Finno-Ugric) groups. This stage, and the previous stage, may be labelled SECONDARY HOMELAND, which was larger, and included a more eastern area.

Romanoslavica XLIII

93

– During this gradual move to south, later also to east and west, the Proto-Slavic nucleus encountered various ethnic groups; they cohabitated with some of them (the Avars), with others, for sure, had more or less important conflicts.

– When the Byzantines first met the Sclavini, it is doubtful that we may speak of a ‘pure’ Slavic idiom; at that time, the linguistic coagulation was still in full process, and – beside the Proto-Slavic nucleus, there surely were still unassimilated Thracian and West Iranic groups, as well as ethnic groups of various other origins, including Altaic groups. Some Thracian groups must have been somewhat isolated, and did not participate in this ample process of linguistic coagulation. Some of them were later assimilated, some are the ancestors of the Albanians (the Shqiptari). Albanian is a Neo-Thracian (not a Neo-Illyrian) idiom, with a specific structure and an important Romance vocabulary; some of this vocabulary is shared with both Romanian and West Romance languages, but some is specific, the best proof that, some time after the ‘Sclavenic period’, Romanians and Albanians gradually became isolated in linguistic islands. There has never been a complete insulation of these ethnic groups, and some Romanian linguistic islands still survive until now in the South Danubian regions of the Balkans.

– Towards the half of the 9th century A.D., the process of what we may label the Slavic ethnogenesis seems to have reached its final phase; the missions of Cyrill and Method supported this process by creating a literary language. At the same time, the new historical and political context led to the emergence of new ethnic groups as always in Europe. I assume that the conscience of a common heritage, of slovanstvo, was gradually reached at during the 6th through the 11th centuries, a complex process, which accompanied the administrative, political and social organisation of the Slavic groups.

– The process of ‘Slavic Making’, to use Curta’s term, was parallel to other more or less similar processes, both on the ruins of the Roman Empire, and in Barbaricum. From this point of view, disregarding whether we use the standard formula ‘Slavic ethnogenesis’ or ‘Slavic making’, it was a process parallel with other similar processes of ‘ethnicisation’ all over Europe. None of these ‘ethnic processes’, including the gradual conscience of belonging to a certain linguistic and religious group, was identical to another. There were identical, similar or different data, specific to every ethnic group. It is hardly believable that we may ever have an ‘ethnic rule’ for those times, from – say –

Romanoslavica XLIII

94

5th to 10th century A.D. The social and military context, the relations with their neighbours, the rule of the most powerful, and hazard played their role.

Conclusions

The Slavic ethnogenesis (or ‘making’) was the result of three basic satem

amalgamation: South Baltic (conventionally ‘Proto-Slavic A’), West Iranic (‘Proto-Slavic B’), North Thracian (‘Proto-Slavic C’), with Germanic and early East Romance (Proto-Romanian) elements. My reconstruction is that this process of amalgamation of the three A-B-C satem elements began some time after the 4th century A.D., and continued ‘in move’ (as Godłowski assumes) for about five centuries. A final phase, the literary coagulation, began some time after 860 and continued in the long and complex process of emerging new ethnika.

We have all the reasons to assume that the first satem groups beginning their expansion in the 6th century A.D. did not have a consistent language, but rather spoke more or less related satem idioms, some of them definitely spoke languages belonging to other linguistic families. In the long run, the three main A-B-C satem groups merged into a more consistent and congruent ethno-linguistic structure to be later known as Slavic. For sure, the term Sclavus circulated at colloquial level, then the forms Sclavenus, Sclavinus, pl. Sclaveni, Sclavini gradually became common in the Byzantine documents. The origin of Sclavus, hence Romanian șchiau, pl. șchei ‘Slav(s)’, may be debatable, seemingly was deformed and/or adapted from a form derived from Slověninъ, pl. Slověne. Disregarding the ultimate origin, it is quite clear that the form Sclavus, Sclavenus, Sclavinus did not initially have an ethnic meaning, at least not in the sense we are accustomed to use the term ethnonym. It rather had social and military meanings, to a less extent a linguistic and scientific meaning as we should expect. The same may be stated for the Arabic borrowing Ṣaqlab (Ṣiqlab, Ṣaqlāb), pl. Ṣaqāliba, behind which we may find people belonging to completely different ethnic groups, and whose common denominator was ‘blond Slave, a Slave with white skin’.

The amalgamated character of these groups is also proved by the same origin of the Albanian forms derived from the same form Sclavus, i.e. Shqip ‘Albanian’ (adj.), Shqiptar (*sklya-b-); Shqinikë < Sclavenica (Dardania, i.e. regio sclavenica); Shqa, Shkla, Shkle ‘a Bulgarian’; these forms also suggest that sparse, non-Romanised Thracian groups contributed to the Slavic ethnogenesis, and also represented an important component of the Albanian ethnogenesis:

Romanoslavica XLIII

95

moving southwards, some of them merged with other satem speakers to eventually become the Sclavini, and other groups moved south-west and, in amalgamation with the Dalmatian (formerly Illyrian) Romanised population led to the Albanian ethnogenesis. I am inclined to consider Albanian a neo-Thracian, rather than neo-Illyrian idiom, even if the Illyrian tradition was locally preserved, and some forms – mainly place-names – were later adapted to the new, emerging idiom later known as Albanian, or gjuha shqipë. In North Danubian regions, the North Thracian groups known as Daci Liberi (Free Dacians) were later assimilated by the already Romanised Thracian groups of the first phase after the Roman conquest. It is probable that Thracian speakers survived in both North and South Danubian areas until at least the 6th century A.D., if not even later. Archaeologically they may be identified until the 7th century A.D., but their survival may be postulated even later.

By the 10th century A.D., this long process of amalgamation and ethnic changes was basically concluded, and the new Slavic groups began their new history in the new Christian context. The Slavic ethnogenesis did not essentialy difer from other similar, but not identical, complex processes. For sure, the centurylong Slavic expansion and ethno-linguistic consolidation was too vast and complex to be fully presented here, but the main issues have been hopefully approached.

Romanoslavica XLIII

96

Romanoslavica XLIII

97

Romanoslavica XLIII

98

Romanoslavica XLIII

99